Two engineers, seven builds, no sales team.
Assiz is small on purpose. Both founders write code, both talk to clients, and nothing goes on this website that isn't running somewhere for someone.
Who we are
Assiz Ltd is a UK-registered AI company working across healthcare, pharmaceutical research, nonprofit infrastructure and education. Two engineers founded it and two engineers still run it. There is no layer between the person you talk to and the person who builds the thing.
The work covers clinical documentation for physicians, evidence-tiered research retrieval for pharmaceutical teams, sponsor platforms for nonprofits, and applied AI teaching in schools. On paper those look unrelated. In practice they share one property: in every one of them, software that sounds right while being wrong causes more harm than software that admits it doesn't know.
That is the whole company, really. We spend most of our engineering effort on the refusals.
Three habits we don't negotiate on
Production from day one
Secure API routing, role-based access and audit trails go in during the first sprint. Retrofitting them after a successful pilot is how good projects turn into expensive ones.
Collect as little as possible
Every build opens from a data-minimisation position. A system gets what it needs to function and nothing beyond that. The Builders Lab collects nothing at all, which took work.
We stay after handover
Handing over a repository and disappearing is easy and we don't do it. Assiz stays through deployment and keeps iterating while the organisation moves around the system.
The two of us
We have worked on every product in this portfolio together, from the first commit forward.
Areeb Husain Arastu
Areeb leads technical architecture and client delivery across the Assiz portfolio. His route into AI ran through clinical research rather than around it. During NHS clinical placements at Chelsea and Westminster Hospital he ran a quality improvement project on the Care of the Elderly ward, developing a delirium screening protocol that improved early identification rates by 30% and presenting the findings at the British Geriatrics Society Spring Meeting and the European Geriatric Medicine Society Congress.
His published work spans geriatric medicine, cardiology and evidence synthesis, including a co-authored time trend analysis of call-to-needle times for STEMI patients accepted at the American College of Cardiology 2026 Scientific Session, and meta-analyses on prenatal alcohol exposure and vitamin D in NAFLD. That grounding shapes how Assiz builds for healthcare: the constraints on a clinical system are ones he has worked inside, not read about.
PGDip International Medical Sciences (Merit), University of Central Lancashire · BA Magna Cum Laude, Fairmont State University · Research presented at BGS, EuGMS and ACC · ICH Good Clinical Practice E6(R3) certified
Zoha Fatema Mirza
Zoha leads technical development and operations at Assiz. A computer science specialist and creator of voice-enabled AI tools, she built Hop, the live sponsor-tracking platform serving Haven of Peace in Tanzania, and has contributed substantially to every Assiz product since the beginning. She brings an unusually human lens to systems design, asking who a piece of software actually has to serve before asking how it should be built.
Having spent most of her life in East Africa, and travelled extensively across India, the Middle East and the United States, she brings a genuinely global perspective to products intended to work in very different contexts.
Microsoft Certified agentic developer · BSc Computer Science, University of the People (California), full scholarship
You'd be talking to the people who build it.
A first conversation costs nothing and commits you to nothing. Worst case, you get a second opinion for free.